Introduction

Previous reports have shown that among ST elevation acute myocardial infarction (STEMI) patients, the development of acute kidney injury (AKI) occurs during hospitalization in 10–27 % of acute MI patients and predicts both short- and long-term mortality [1–5]. Since no standardized definition of AKI has been used previously for patients with MI, information on the true incidence as well as the clinical and prognostic are still lacking. The recently proposed AKI network (AKIN) [6] provides a standardized definition of AKI, and includes three stages based on the changes in serum creatinine. Although the use of this classification was shown to predict mortality in a variety of critically ill patients including septic, post cardiac surgery and following angiography [7–10], limited information is available regarding the use of the AKIN classification to evaluate the incidence and clinical relevance of AKI among STEMI patients, especially in those undergoing primary percutaneous intervention (PCI), as well as its relation to long-term survival. In the present study, we evaluated the incidence, risk factors, in-hospital complications as well as the short- and long-term mortality associated with AKI as classified by AKIN definition in a large single-center cohort of consecutive STEMI patients.

Methods

Study population

A retrospective, single-center observational study was conducted in the Tel-Aviv Sourasky Medical Center, a tertiary referral hospital with a 24/7 primary PCI service. Included were all 1,105 consecutive patients admitted between January 2008 and November 2012 to the cardiac intensive care unit (CICU) with the diagnosis of acute STEMI. Excluded were 18 patients who were treated either conservatively or with thrombolysis, and 45 patients whose final diagnosis on discharge was other than anterior wall STEMI (e.g. myocarditis or Takotusubo cardiomyopathy). Also excluded were patients who died within 24 h of admission (n = 6) because they would not have had sufficient time to develop AKI and patients requiring chronic peritoneal or hemodialysis (n = 3) treatment. The final study population included 1,033 patients (Fig. 1) whose baseline demographic, cardiovascular history, clinical risk factors, treatment characteristics and laboratory results were retrieved from their medical files.

Fig. 1
figure 1

Study flow chart

Full size image

Protocol

The diagnosis of STEMI was established by a typical history of chest pain, diagnostic electrocardiographic changes, and serial elevation of serum cardiac biomarkers [11].

Primary PCI was performed in patients with symptoms ≤12 h in duration as well as in patients with symptoms lasting 12–24 h in duration if the symptoms continued to persist at the time of admission. Following primary PCI, left ventricular ejection fraction was measured in all patients, by bedside echocardiography, within the first 48 h of admission. Patient records were evaluated for in-hospital mortality and complications occurring during the hospitalization. These included cardiogenic shock or the need for intra aortic balloon counterpulsation (IABC) treatment, need for emergent coronary artery bypass graft (CABG) surgery, mechanical ventilation or heart failure episodes treated conservatively, clinically significant tachyarrhythmias (ventricular fibrillation, sustained ventricular tachycardia, and atrial fibrillation) and bradyarrhythmias requiring pacemaker as well as major bleeding (requiring blood transfusion). Mortality was assessed over a median period of 1,526 ± 298 days (range 2–2,130) up to August 1, 2013. Assessment of survival following hospital discharge was determined from computerized records of the population registry bureau. The study protocol was approved by the local institutional ethics committee.

Laboratory

The serum creatinine (sCr) was determined upon hospital admission and at least once a day during the CICU stay and was available for all analyzed patients. The estimated glomerular filtration rate was estimated using the abbreviated modification of diet in renal disease equation (MDRD) [12]. Baseline renal insufficiency was categorized as an estimated glomerular filtration rate (eGFR) of ≤60 ml/min/1.73 m2 [13]. AKI was defined, applying the AKIN classification [6], according to the maximum increase in sCr from baseline (hospital admission): stage 1—an increase of ≥0.3 mg/dl in sCr; stage 2—more than two- to threefold sCr increase; stage 3—more than threefold sCr increase from baseline or baseline sCr >4.0 mg/dl with an acute increase of >0.5 mg/dl; or the need for renal replacement therapy, irrespective of the AKIN stage at renal replacement therapy initiation. Renal recovery was defined as a reduction in sCr to the admission range without the need for renal replacement therapy.

Statistical analysis

All data were summarized and displayed as mean (±standard deviation) or median (25–75 %) for continuous variables and as number (percentage) of patients in each group for categorical variables. The p values for the Chi-square test were calculated with the Fisher’s exact test. Continuous variables were compared using the independent sample t test or Mann–Whitney test. The identification of the independent predictors of AKI was assessed using logistic regression. Binary logistic regression models were performed using the Enter mode. The models were adjusted for age, male gender, left ventricular ejection fraction, hypertension, diabetes mellitus, eGFR and symptom duration. The influence of AKIN on the occurrence of all-cause mortality was evaluated using multivariate Cox regression. We adjusted for age, gender, hypertension, diabetes mellitus, smoking status, left ventricular ejection fraction, eGFR, peak creatine phosphkinase (CPK),baseline hemoglobin, white blood cell count, high sensitive C reactive protein (CRP) levels, and AKIN status. A two-tailed p value <0.05 was considered significant for all analyses. All analyses were performed with the SPSS 20.0 software (SPSS Inc., Chicago, IL).

Results

A total of 1,033 STEMI patients treated by primary PCI were enrolled in the study, 100 (9.6 %) of whom developed AKI in accordance with the AKIN criteria. Stage 1 AKI occurred in 79 patients (79 %, mean sCr increase 0.46 ± 0.21 mg/dl), AKI stage 2 occurred in 14 patients (14 %, mean sCr increase 1.3 ± 0.29 mg/dl) and AKI stage 3 developed in 7 patients (7 %, mean sCr increase 2.26 ± 1.45 mg/dl). Only 2 patients among those developing AKI required renal replacement therapy throughout hospitalization, whereas 57 (57 %) of the AKI patients had renal recovery at CICU discharge. A total of 257 patients (25 %) had baseline renal insufficiency upon hospital admission.

The baseline clinical characteristics of patients with and without AKIN are listed in Table 1. Patients with AKI were more likely to be older, of female gender, to have more co-morbidities, longer symptom duration prior to emergency room admission, lower baseline eGFR, lower left ventricular ejection fraction and longer time until hospital discharge.

Table 1 Baseline characteristics
Full size table

The occurrence of AKI following STEMI treated with primary PCI resulted in more complications and adverse events during hospitalization, as well as higher 30-day mortality (Table 2).

Table 2 In-hospital complications
Full size table

Long-term outcome

Over a mean period of 2.7 + 1.6 years, 72 (6.9 %) patients of the entire cohort died. Mortality was significantly higher among those with AKI (29/100, 29 %) following STEMI than those without AKI (43/933, 4.6 %); p < 0.001). Among the 29 patients with AKI who died, 20 (69 %) had AKIN stage 1 and 9 patients (31 %) had AKIN stages 2 and 3. Figure 2 shows the Kaplan–Meier survival curve for long-term survival according to the AKIN stages. A significant increase in long-term mortality was observed between patients without AKI and those with stage 1 and 2/3 AKIN (p < 0.001). AKI was an independent predictor of mortality reaching a hazard ratio of 6.68 (95 % CI 2.1–21.6, p = 0.002) compared to patients without AKI. Within the AKIN groups the hazard ratio for all-cause mortality was 7.42 (95 % CI 1.9–27.8, p = 0.003) for stage 1 AKIN, and 5.55 (95 % CI 1.1–28.8, p = 0.04) for combined AKIN stages 2 and 3.

Fig. 2
figure 2

Cumulative survival rates for 1,033 patients with ST elevation myocardial infarction on the basis of AKIN (acute kidney injury network) severity classification

Full size image

Predictors of AKI

We created a multivariate regression model in order to elucidate the risk factors predicting the development of AKI. The major risk factors found for AKI occurrence were hypertension (OR 2.6, 95 % CI 1.5–4.3, p < 0.001), gradual 1 % reductions in left ventricular ejection fraction (OR 0.95, 95 % CI 0.92–0.97, p < 0.001) and eGFR (OR 0.972, 95 % CI 0.95–0.99, p = 0.005) and a trend for every 1 year of age increment (OR 1.025, 95 % CI 0.96–1.05, p = 0.09) (Table 3).

Table 3 Binary logistic regression model for predicting AKIN
Full size table

Discussion

The main finding in our study is that early AKI is a frequent complication, occurring among 9.6 % of STEMI patients undergoing primary PCI, and that even mild elevation of sCr (≥0.3 mg/dl, AKIN stage 1) is associated with a marked increase in both in-hospital and all-cause mortality, up to 5 years following STEMI. To our knowledge, this is the first study assessing both short- and long-term mortality in a large non-selected population of patients presenting with STEMI and treated by primary PCI, in whom the AKIN classification was assessed following daily measurement of sCr.

The worsening of renal function throughout hospitalization in STEMI patients is multifactorial and is affected by hemodynamic state, contrast material use during cardiac catheterization, drugs admitted (especially blockers of the renin angiotensin axis) as well as the occurrence of sepsis, bleeding, atheroembolic disease and acute hyperglycemia [14–16].

The lack of consensus criteria to define AKI occurrence and severity had resulted in marked variability of reported AKI among AMI patients ranging from 10 to 27 %, depending on the study population considered, the period during which the study was conducted and, in particular, the criteria used to define AKI [1–5]. Several definitions of AKI have been used in AMI studies. Some studies defined AKI as an absolute (≥0.3 or ≥0.5 mg/dl) or relative (>25 %) increases in sCr above baseline at any point during hospital stay, without stratification of AKI severity. Other studies proposed a staged classification of AKI based on the both absolute and relative sCr changes [17, 18]. Moreover, sCr was evaluated at different time frames, ranging from 24 h following AMI to 1 week [1, 3, 10, 14, 15]. The different definitions of AKI and heterogeneous data reported have lead to difficulties in comparing the various study results. In order to amend the lack of a standardized definition, the Acute Dialysis Quality Initiative Group (ADQI) proposed in 2002 the classification Risk, Injury, Failure, Loss and End-stage Renal Failure (RIFLE) classification [19, 20]. The AKIN proposed in 2005 improvements to the RIFLE criteria in order to make the detection of AKI swift and clear, by adopting a lower cutoff point for variation in serum creatinine [6]. Although the importance and prognostic implications of the AKIN classification were validated in various critical care conditions [7–9, 21–23], limited data exists regarding the utilization of AKIN in ischemic patients.

Hwang et al. [24] reported a 30 % prevalence of AKI, defined by AKIN among AMI patients which independently predicted 1-year mortality, with stage 3 AKIN STEMI patients having the worst outcomes. No information was reported on the distribution of AKIN between STEMI and non-STEMI patients, and regarding the use of PCI. Marenzi et al. [25] evaluated the incidence and outcomes associated with AKI, defined by AKIN, in a large group of acute coronary syndrome patients. In that report AKI prevalence was 13 % (14 % among STEMI patients) and associated with a more complicated in-hospital course and higher in-hospital mortality. AKI was more prevalent among STEMI patients, and those undergoing primary PCI; however the latter was performed in only 82 % of STEMI patients. Our study is novel in several aspects, since it is the first to evaluate AKI and long-term all-cause mortality in a large homogeneous cohort of STEMI patients, all undergoing primary PCI. The lower incidence of AKI in our cohort (9.6 % compared to 14 % in the STEMI group described by Marenzi [25]), although a higher rate of primary PCI in our cohort, may theoretically point out that early PCI in STEMI patients may protect against AKI development, as it allows early normalization of cardiovascular hemodynamics. Although our findings demonstrated a lower hazard ratio for long-term mortality among AKIN stages 2 and 3 patients as compared to those with AKIN stage 1, we believe that these results reflect the lower number of patients having AKIN 2/3 (21/1,033, 2 %), and that eventually a higher AKIN stage will be associated with a higher risk for mortality.

Our findings bare some important clinical implications. The close association between even a small elevation (≥0.3 mg/dl) in sCr and increased morbidity and mortality observed in our study highlights the importance of obtaining a baseline sCr value early after admission for STEMI, and the need for frequent monitoring of renal function during the entire hospitalization period. The identification of AKI predictors, such as hypertension, age, baseline eGFR, and left ventricular ejection fraction may help in pointing those at higher risk, warranting even more frequent monitoring and prompt intervention. In those patients the early implementation of prophylactic strategies aimed at preventing AKI or at reducing its severity might also provide significant clinical benefit. Recent trials demonstrated that among patients with chronic renal disease a single high loading dose of atorvastatin administered within 24 h before contrast material exposure is effective in reducing the rate of contrast-induced AKI [26, 27]. In addition, rapid alkalization by bolus injection of sodium bicarbonate was effective for the prevention of contrast-induced nephropathy (CIN) in patients with chronic kidney disease undergoing emergent procedures [28].

Another novel treatment strategy recently shown to decrease contrast-induced AKI is remote ischemic preconditioning (rIPC), defined as transient brief episodes of ischemia at a remote site, and represents an adaptive response that protects against ischemic and reperfusion insult in various organs [29, 30]. Studies demonstrated that rIPC has protective effects on renal function among patients undergoing elective PCI [31, 32] and was shown to increase myocardial salvage among STEMI patients undergoing primary PCI [33]. In this regard, rIPC may offer a novel noninvasive and virtually cost-free treatment strategy for improving cardiovascular hemodynamics and decreasing AKI incidence in this population.

Finally, as AKI was associated with higher long-term mortality, it should be regularly reported as part of the in-hospital course, even if mild and transient.

The major limitation of our study was lack of information regarding the volume of contrast material used during primary PCI, which was available only for a minority of patients. Thus, the effect of contrast volume on renal impairment was not included in the analysis. Contrast-induced acute kidney injury (CI-AKI) is a prevalent and deleterious complication of coronary angiography and reported to be the third most common cause of hospital-acquired renal failure [34]. The risk of CI-AKI is directly associated with increasing contrast media volume [35], with incidence ranging from 2 % among patients with normal baseline renal function to as high as 20–30 % in patients with a baseline creatinine >2.0 mg/dl [36]. Even after adjusting for baseline renal function and co-morbidities, in-hospital mortality is about fivefold higher in CI-AKI patients, and 1- and 5-year mortality rates are about fourfold higher [37]. We believe, however, that additional factors aside from CI-AKI contribute to AKI development in this setting [14–16]. Our above-mentioned results show that patients developing AKI had longer symptom duration and worse LV function, which can alter hemodynamics and renal perfusion. Symptom duration in AMI is a powerful prognostic marker in AMI patients undergoing reperfusion [38, 39] and major consideration is given to minimizing ischemic duration in order to improve survival following an AMI [40]. Time to reperfusion is known to determine infarct size [41, 42] and systolic function [42, 43], which was also true in our cohort. In addition, we recently demonstrated that longer symptom duration is associated with higher admission C reactive protein and lower hemoglobin level in STEMI patients [44]. It appears thus that longer symptom duration may alter left ventricular function and hemodynamics as a consequence of inflammation, thus contributing to AKI development.

Our study bares some other notable limitations. First, this was a single-center retrospective, non-randomized observational study, and as such may have been subject to bias, even though we included consecutive patients and attempted to adjust for confounding factors using the multivariate Cox regression model. Second, although AKI definition refers to a sCr increase compared to the baseline value, the sCr at hospital admission may not represent a true baseline value in STEMI patients as an increase could have already occurred prior to hospital arrival owing to the hemodynamic impairment. Third, data regarding concomitant therapy with statins, Renin/angiotensin blockers and diuretics throughout hospitalization were not present for many patients, and their effect on AKI development could not be assessed. Finally, the definition of AKIN refers to sCr change within a time frame of 48 h. As the change in sCr can lag beyond this time period due to delayed effects of contrast material and drugs, worsening of renal function might have occurred following hospital discharge in some patients, thus the true incidence of AKI described in our study may have been an underestimation.

We conclude that AKI defined by AKIN is a frequent and hazardous complication in STEMI patients in the primary PCI era, and its occurrence is associated with adverse short- and long-term outcomes.